Development of Active Wind Vane for Low-Power Wind Turbines

Abstract

This paper proposes the development of an active control system to control the power output of a low-power horizontal-axis wind turbine (HAWT) when operating at wind speeds above the rated wind speed. The system is composed of an active articulated vane (AAV) in charge of the orientation of the wind turbine, which is driven by an electric actuator that changes the angle of the AAV to maintain a constant power output. Compared with the passive power regulation systems most often used in low-power HAWTs, active systems allow for better control and, therefore, greater stability of the delivered power, which reduces the structural stresses and allows for controlled braking in any wind condition or during system failures. The control system was designed and simulated using MATLAB R2022b software, and then built and evaluated under laboratory conditions. For the control design, the transfer function (TF) between the pulse width modulation (PWM) and the AAV angle ($\theta$) was determined via laboratory tests using MATLAB’s PIDTurner tool. For the simulation, the relationship between the power output and the AAV angle was determined using the vector decomposition of the wind speed and wind rotor area. Wind speed step and ramp response tests were performed for proportional–integral–derivative (PID) control. The results obtained demonstrate the technical feasibility of this type of control, obtaining settling times (ts) of 6.7 s in the step response and 2.8 s in the ramp response.

1. Introduction

During the last decade, wind energy has experienced significant growth [1]. In particular, there has been a notable increase in the size and power of HAWTs [2]. This evolution raises the question of the difference between a standard HAWT and a low-power HAWT. Low-power HAWTs are designed to generate smaller amounts of energy compared with conventional HAWTs, which raises interest in understanding their specific characteristics and applications. These systems may be more suitable for certain environments and applications where a more limited but equally crucial power generation is required, such as in rural areas or remote communities.

There are obstacles that low-power HAWTs must overcome in the domestic renewable energy sector. This challenge is due to the relatively low wind velocities in most urban areas and the continuing decline in the cost of photovoltaic electricity. As a result, low-power HAWT installation is limited to areas with average wind speeds of more than 5 m/s. The creation of innovative urban wind turbine designs that have qualities like cheap cost per kWh produced, plenty of surface area to capture even low wind speeds, silent operation, and animal friendliness is required to address this scenario. Wind-direction-orientable HAWTs typically outperform those without this feature in terms of efficiency. HAWTs can also be made more versatile by facing away from the wind by using a rear rudder. Thus, the safe functioning of low-power HAWTs depends on the development of efficient braking systems. To avoid harming surrounding living things or machinery, this becomes even more important. Numerous braking systems have been created in this area, mainly falling into three categories: electrical, mechanical, and aerodynamic [3].

The problem of frequent wear affects mechanical brakes, which are commonly used in HAWTs. This might result in higher maintenance costs or the need to replace components. Disk brakes are typically preferred when friction is used to stop the HAWT rotor. Mechanical braking devices are useful, but larger HAWTs also need them for maintenance or as a safety precaution in bad weather. Aerodynamic braking is an additional option that reduces the lift force by modifying the blade angle in reaction to severe winds. This approach gives the user more control over the system, but it also makes the HAWT structure heavier. Aerodynamic brakes are a good alternative for high-power systems since they may be used repeatedly without losing their effectiveness, despite their initial expensive cost.

Figure 1 shows a characteristic power curve of a HAWT, sectioned into four regions. In Region I, the wind turbine is stopped by low wind speeds (the control is inactive). In Region II, the wind turbine operates under maximum power tracker control. In Regions III and IV, constant power regulation control operates over the entire range and rotor braking, respectively [4].

HAWTs are typically controlled to align the rotor plane perpendicular to the wind direction, ensuring maximum energy transformation [5]. This strategy increases the energy production; however, high wind loads and centrifugal forces generate stresses that damage the structure. To reduce this damage and maximize the efficiency, new control, regulation, and guidance systems have been developed that continuously monitor and adjust the rotor position in response to changing wind conditions, contributing to efficient and safe system operation.

1.1. Orientation System

In low-power wind turbines, regulation and stop systems have been implemented according to the IEC-61400-2 standard [6], which sets the standards for such systems, including criteria and methodologies to ensure their safety, reliability, and effectiveness. In these low-power wind turbines, soft-stop systems have been implemented for rotor orientation and compared with other orientation strategies through steady-state analysis and dynamic simulation [7]. Additionally, passive orientation mechanisms with a fixed pitch angle have been presented, including the tilting articulated vane and offset pivot systems, which are simple and economical [8]. The IEC-61400-2 provides detailed guidelines on specific aspects of regulation and control systems, such as the response to different wind conditions, protection against overloads and emergency situations, interoperability with other components of the wind energy system, and maintenance and compliance requirements.

1.2. Braking Systems

The braking system of a wind turbine must maintain power production under given operating conditions; it must also be economically viable and reliable. To achieve this goal, researchers in [9] developed a hydraulic brake for a 100 kW HAWT with a braking torque of 1.49 kNm. They estimated that under real conditions, the rotor will stop in 1.1 s. Researchers in [10] devised a dynamic braking system with a controllable resistive load connected to the generator and activated using active power control and a switch, preventing the system from coasting. In [11], the authors gently reduced the rotational speed of the HAWT using a Y-circuit with negative temperature coefficient thermistors connected directly to the generator terminals. Additionally, ref. [12] implemented a braking system for a small wind turbine prototype using the DRV8873 H-bridge motor driver. This controller allows for bidirectional control of the DC motor with a peak current of up to 10 A, providing integrated protections and an interface for mechanical end stop switches. During testing, the braking system proved effective and reliable, safely stopping the wind turbine rotor in the case of an emergency.

In [13], an innovative method was introduced for controlling small-scale vertical-axis wind turbines with a magnetic brake in two modes. One mode gradually slowed down the turbine when it reached its maximum speed, while the other maximized energy extraction by adjusting a coefficient. The proportional controllers in the simulations met the requirements and reference inputs. Additionally, a supervisory control algorithm and a methodology for its implementation are presented.

1.3. Regulating Systems

The power regulation in a HAWT (Region III of Figure 1) is designed to control the stability of the power being extracted from the wind and to prevent damage to the generator. In addition, it guarantees the constant and safe delivery of electricity to the grid above the rated speeds.

There are two types of control systems for regulating wind turbine power: passive systems and active systems. The latter require high torque actuators to move the generator mass, resulting in slow movements and high maintenance costs [14], unlike passive systems that act through the force exerted by the wind.

Active control systems for small wind turbines regulated the rotational speed and kept the power constant in the rated region. Notably, the YAW and PITCH control systems were related to the electrical production efficiency [15]; the YAW systems regulated the power in regions higher than nominal speeds, using the On/Off technique, and proposed an orientation speed of 3 °/s as the most appropriate to limit the power [16]. On the other hand, YAW systems presented an error defined by the variation between the wind direction and the nacelle position, which caused the generated power to decrease [17].

For the design of the YAW control, models based on HAWT output power [18,19,20,21] were used, where the power coefficient model (Cp) was an essential element to determine the system power. Additionally, there were models based on vector decomposition to vary the angle of attack [16,22,23,24] or the catchment area [25,26,27,28,29,30,31], which were parameters that determined the power loss [16].

1.4. Control Techniques

Control techniques in HAWTs improve the power stability above the rated speeds. These techniques adjust the power with respect to the wind and other environmental factors, maximizing the power generation and ensuring a stable electricity supply.

In high-power systems, there were advanced and intelligent techniques that were shown to be effective, such as On/Off control, PID, PI, fuzzy logic, and PID–fuzzy logic [32,33,34,35,36,37,38], where PID or fuzzy logic control had similar and effective responses, with PID being a simpler control structure. However, these techniques were not extended to low-power HAWTs. It was also identified that regulation systems, such as PITCH and YAW, coped with speeds that were higher than nominal; vane systems took advantage of the wind load to drive the mechanism, allowing for the control and braking of the devices, and were economical and simple mechanisms.

Therefore, the main objective of this work was to develop a power regulation and braking system with an active vane for low-power HAWTs, highlighting its economical and effective nature. To achieve this, a PID control technique was implemented that offers the advantage of maximizing the utilization of available wind resources. The implementation of these systems allows for a significant improvement in the energy efficiency of the wind turbine, ensuring optimal performance, even under variable wind conditions. Furthermore, the use of PID control ensures stability in power generation, which contributes to maintaining a constant and reliable supply of electrical energy. Additionally, these systems help to protect both the wind turbine and other system elements against potential damage due to abrupt fluctuations in wind conditions, thus ensuring safer and more reliable operation over time.

This document is structured as follows: Section 2 presents the design of the proposed AAV system with the design characteristics and the block diagram used to simulate the control of the AAV system in MATLAB Simulink. Section 3 provides details on the simulation of the AAV system implemented in an RTO ENERGY 1000 W wind turbine model for two wind inputs, and in Section 4, a discussion of the results obtained regarding other methods used for the regulation and braking of a low-power HAWT is presented.

2. Materials and Methods

A low-power HAWT was used with the characteristics presented in

Table 1. Table of parameters for the RTO ENERGY 1000 W wind turbine [39].

Parameter	Variable	Value	Unit
Nominal electrical power	$P_{\mathrm{rated}}$	1000	W
Reference electrical power	$P_{\mathrm{ref}}$	1000	W
Cut-in wind speed	$V_{\mathrm{cut}}$	13	m/s
Rated wind speed	$V_{\mathrm{rated}}$	11	m/s
Start-up wind speed	$V_{\mathrm{in}}$	3	m/s
Mechanical limits	$\theta$	0–72	°
Diameter of the rotor	$D_R$	2	m
Rotor blade inertia	$R_i$	0.8141	kg·m2
Wind turbine speed at rated wind speed	$R_s$	950	rpm
Electrical generator lumped resistance	$G_R$	0.616	Ohm
Electrical generator lumped leakage inductance	$G_I$	0.000277832	H
Electrical generator effficiency		90

. The wind turbine was designed to remain parallel to the wind direction in Regions I and II, to maintain constant power at higher wind speeds (Region III) by regulating the AAV angle, and to execute smooth and safe braking in an emergency (Region IV) by placing the AAV perpendicular to the wind direction.

Figure 2 illustrates the mechanical scheme of the wind turbine with the AAV. This mechanism was operated by an actuator, which controlled the YAW angle and the AAV angle ($\theta$), with the latter being formed between the plane perpendicular to the wind rotor and the wind direction.

The angle ($\theta$) influenced both the wind vector decomposition and the area vector decomposition, thus impacting the energy capture efficiency of the wind turbine. Taking this into account, the power of the wind turbine can be expressed by Equation (1), where the YAW factor is given by (${cos}^4\theta$).

$$P=\frac{1}{2}\rho A{C}_p{V}^3{cos}^4\left(\theta \right)$$

(1)

2.1. Design and Construction

From the model proposed in Figure 2, the mathematical model of the power curve was estimated. Taking into account the variation in the YAW angle, the mechanical design of the AAV system represented in Figure 3, which was built with the mechanical and electrical parameters of the actuator from

Table 2. Table of parameters for the AAV in an RTO ENERGY wind turbine.

Parameter	Value	Unit
Vane area	0.13	m2
Rudder length a	1146.07	mm
Rudder length b	460.37	mm
Blade radius	951.76	mm
Actuator torque	100	N
Actuator speed	90	mm/s

, was carried out. Subsequently, the prototype of the system was developed with the proposed mechanical considerations, with the aim to obtain the transfer function of the design, which allowed for analyzing its dynamic behavior and responses to different operating conditions.

Once the prototype was obtained, a step test was performed with the design parameters, which used a PWM from 0 to 255 bits as the input and the angular movement of the AAV as the output. To obtain the transfer function, a position sensor placed on the axis of motion of the AAV was used to determine the position angle, which employed the “system identification” method [40], and the data obtained were collected in arrays with sampling times of 0.038 ms.

Subsequently, the signals and statistical analysis techniques were utilized (according to the time mentioned in the previous paragraph) to identify patterns and trends in the data in relation to the difference between the measured power ($P_{\mathrm{feed}}$) and the reference power ($P_{\mathrm{ref}}$), called the input power ($P_{\mathrm{in}}$), which was proportional to the PWM. This was achieved using the PIDTurner tool in MATLAB. Thus, after obtaining the TF of the AAV, together with the diagram in Figure 4, the step and ramp tests were performed under wind variation.

The tests were performed in MATLAB Simulink with a step input and wind ramp to determine the control parameters: time ($t_s$) and overshoot ($M_p$).

2.2. Wind Turbine Model

First, the system was defined using blocks that represented each component of the control process. It started with a measurement block to capture the output power of the wind turbine and calculated the error by subtracting this output power from the desired reference power.

Figure 5 depicts the block diagram of the implemented PID control [41] to calculate the pulse width modulation (PWM) value that ensured a position of $\theta$ for power stability. The term “minimum PWM” denotes the minimum output value assigned to the PID controller that ensured the system response initiation. The proportional term provided feedback based on the current error, while the differential term offered a smooth system response to high-frequency error fluctuations. Additionally, the integral term facilitated the correction of steady-state error.

Then, a PID block was used to determine the control signal that would correct the calculated error. This control signal was applied through a PWM block, which simulated the control action on the actuator speed.

The next block represents the transfer function (TF), which models the relationship between the wind speed and the AAV angle. This angle was limited by the mechanical limit block of the prototype with the design data from Table 2. To evaluate the wind effect, a wind input block was used to determine the wind speed in the wind turbine model. Finally, a block representing the wind turbine model [42] with the parameters of the RTO ENERGY 1000 W [39] wind turbine from Table 1 was used to estimate the behavior of the output power as a function of the wind speed, which was influenced by the YAW factor that governed the YAW turning behavior.

2.3. Wind Turbine Power Control

A flow diagram of the AAV system control algorithm is presented in Figure 6. This algorithm evaluated the wind speed; when the wind speed exceeded $V_{\mathrm{cut}}$, the control positioned the AAV perpendicular to the flow, which reduced the wind load.

When V was less than $V_{\mathrm{cut}}$ and less than $V_{\mathrm{rated}}$, the controller positioned the AAV parallel to the wind flow. Otherwise, when V was greater than $V_{\mathrm{rated}}$, the AAV system controlled the position of $\theta$ using the value of the input power, which allowed the AAV position to be corrected. When the error was equal to 0, the system stopped; if the error was greater than 0, it rotated clockwise and estimated the actuator speed with a PID controller. If the error was less than 0, the direction of rotation of the AAV changed to counterclockwise and stayed in a range between 0 and 72°, which was mechanically constrained, as shown in Figure 7.

The PID constants were obtained using the method described by Ziegler Nichols and Tyreus Luyben [43], which involved adjusting the proportional gain until the system oscillated periodically. Finally, the critical time was determined and the PID control constants $K_p$, $K_i$, and $K_d$ were calculated.

3. Results

3.1. Design and Construction

The AAV wind turbine system was simulated to evaluate the orientation in Region I, the power limiting in Region III of the power curve, and the braking in Region IV. In addition, the performance of the active guidance strategy was evaluated. The HAWT system parameters are shown in Table 1.

3.2. Transfer Function

By implementing the methodology presented in Figure 4, the PWM step response curve illustrated in Figure 8 was obtained. The red line represents the experimental variation in the angle ($\theta$) that moved from 0 to 78 degrees in 1.6 s, while the black line represents the TF with a coefficient of determination of 91.33%. Equation (2) was approximated from the TF curve, which represents a second-order system with a delay and modeled the experimental response of the AAV system.

$$F_t=\frac{0.02689s+0.30914}{s+0.597372}$$

(2)

By substituting Equation (2) into the model of Figure 7, the block model of the proposed AAV system was obtained.

Using the method of Ziegler Nichols and Tyreus Luyben [43], the PID control constants shown in

Table 3. PID controller constants calculation table.

Constants	Ziegler	Tyreus	Experimental
$K_p$	3	2.27	1.8
$K_i$	0.6	0.103	0.013
$K_d$	3.75	3.607	3.3

were obtained for a critical gain ($K_c$) of 5 and a critical time ($T_c$) of 10 s, as obtained from the AAV block model. By comparing the results of the Ziegler and Tyreus methods, these values were adjusted to obtain the experimental constants.

3.3. Response to the Step

Figure 9a shows a step curve where V changed from 11 m/s to 13 m/s at second 20, causing a rapid increase in the maximum angular velocity ($\omega$). In addition, the limitations on the generator torque due to the reduced current increased the rotor speed sharply. The AAV adjusted the angle to limit the power to 14.66° (see Figure 9c) at a time of 2.1 s before the AAV angle was angled relative to the power. As a result, in Figure 9b, there was an $M_p$ of 39.5% for the power to stabilize at its nominal value, with a $t_s$ of 6.7 s.

3.4. Response to Ramp

A wind speed of 11 m/s to 13 m/s in a ramp-type impulse was applied to the wind turbine system in Figure 9d.

The simulation results show that the AAV angle curve fits $\theta$ at 14.68° in 1.2 s to limit the power at $P_{\mathrm{ref}}$ (see Figure 9f).

In addition, Figure 9e illustrates an $M_p$ of 5% for the power to stabilize at its final value, with a $t_s$ of 2.8 s.

When V changed from 13 m/s to 16 m/s, which was located between $V_{\mathrm{rated}}$ and above $V_{\mathrm{cut}}$, it was observed that at second 44, the mechanism adjusted the angular position when there was an increase in V from seconds 40 to 44. However, when it crossed through the shear velocity, the system slowed down, placing the AAV at 72°.

Specifically, the implementation of a PID controller with the experimental tuning shown in Table 3 also showed satisfactory results for this input.

3.5. Braking Response

The results of a braking simulation are shown in Figure 9a,d. V changed from 13 m/s to 16 m/s, which activated the braking system when the cutting speed was exceeded. In Figure 9c,e, the AAV adjusted the angle to 72° to cancel the power generation at a braking time ($t_b$) of 3.2 s, as seen in Figure 9b,e.

4. Discussion

A comparison between the AAV system of this study and those presented by other authors can be seen in

Table 4. Comparative regulation table.

Criterion			Step Input		Ramp Input	Braking	Unit
	AAV	[16]	[31]	AAV	[17]	[9]
System	AAV	YAW	YAW	AAV	YAW	YAW
Criterion $\theta$	${cos}^4\theta$	${cos}^3\theta$	$cos\theta$	${cos}^4\theta$	$cos\theta$	-
Control technique	PID	On/Off	Fuzzy logic	PID	PI	On/Off
$P_{\mathrm{rated}}$	1	2.1	20	1	10	100	kW
$M_p$	139.5	170	136.85	105	180	-	%
$t_s$	6.7	8.1	300	2.8	10	-	s
$t_b$	3.2	-	-	3.2	-	1.2	s
$\theta$	14.66	33	83.31	14.68	32	-	°
$\omega$	22.5	5	-	22.5	90	-	$\frac{°}{\mathrm{s}}$

. This comparison highlights the superiority of the AAV in terms of the dynamic performance, efficiency, and control complexity.

The PID-controlled performance of the AAV system significantly outperformed systems from other research using On/Off controls and fuzzy logic.

In the step input test, the AAV system presented a notably lower $M_p$ value (139.5%) compared with those obtained by other authors (170%), and a faster settling time ($t_s$) of 6.7 s. The simplicity of the AAV adds an additional component of affordability and efficiency to the solution.

In the ramp test, the AAV maintained a low $M_p$ rate (105%), in contrast to the findings of other authors (180%), and achieved a shorter settling time ($t_s$) (2.8 s vs. 10 s). This confirmed the AAV’s ability to provide quick and efficient responses.

Furthermore, when compared with the work of [9], it was observed that the AAV stopped the rotor in 3.2 s, while a hydraulic brake stopped the rotor in 1.1 s. However, a hydraulic system is more expensive to implement and maintain than an AAV system.

5. Conclusions

An active control system was developed for an AAV, where an electric actuator drives the AAV and defines the orientation of the wind turbine with respect to the wind direction, reliably ensuring proper control of the regulation and braking stages on the power curve.

The AAV enhances the efficiency, stability, and reduction of structural stress in HAWTs compared with current active systems, which utilize very large and expensive final control elements, unlike this new system.

The results of the step and ramp tests demonstrated that the system was capable of maintaining a coefficient of variation with times of 6.7 s and 2.8 s, with $M_p$ values of 39.5% and 5%, respectively, and a braking time of 3.2 s.

The method presented in this work demonstrated efficient results in the simulation; however, for practical application, it will be necessary to dimension the AAV structures according to the desired wind turbine power.